Pingping Fan, Li J, Ye Sun, Xicheng Zhang, Miao Yu
Tea, a nutritionally and culturally vital beverage, demands reliable quality grading and cultivar authentication, particularly for high-value functional varieties. Yet the complex tea matrix and subtle compositional differences among closely related cultivars render on-site identification challenging. While nanopore sensing shows great potential, its application to tea analysis remains unexplored. Herein, we present a rapid single-molecule sensing strategy using phenylboronic acid-modified MspA-90PBA nanopores, achieving accurate tea grading and cultivar discrimination via anthocyanin–catechin dual-component fingerprinting. A total of 10 tea polyphenols (2 anthocyanins and 8 catechins) generate distinct current signals. A machine-learning model reaches 94.2% accuracy for polyphenol identification in complex tea matrixes. This platform precisely discriminates anthocyanin-enriched Zijuan tea from its closely related Ziya tea cultivar and grades Zijuan samples of various quality levels. This high-precision approach provides an effective tool for tea authentication, quality evaluation, and adulteration detection, with broad potential for rapid natural product assessment.